If you grew up with Pokémon, you probably remember the thrill of having a pocket-sized companion that followed you everywhere, reacted to your care, and made each adventure feel less lonely. Today, that idea isn’t just fantasy—it’s becoming reality through social robots.
Having already examined educational robots and edutainment robots, we now turn to another rapidly emerging category within the field of robotics: the social robot. Unlike industrial machines that assemble cars, social robots are designed for something much more human—connection. They are physically present in our environment, allowing us to see, hear, and even touch them; they make eye contact, respond to our voices, and use simple gestures or expressions to create a sense of genuine presence.
Think of PARO, the cuddly robotic seal used to calm dementia patients, or Pepper, the wide-eyed humanoid greeting people in banks and airports. There’s ElliQ, a tabletop companion that chats with older adults at home, Furhat, a multilingual concierge at airports, and Lovot, a rolling ball of affection whose only job is to be loved. These aren’t tools for teaching or heavy lifting—they’re companions built to share space with us.
So where do social robots make the most difference? Around the world, their biggest impact so far has been in elder care and mental health—areas where loneliness and daily support are pressing challenges.
What roles do social robots play in elder care and mental health?
Across Japan, Europe, and the United States, ageing populations and caregiver shortages have put loneliness and routine support in the spotlight. Here, social robots are finding a practical niche: offering companionship, structure, and gentle prompts—without pretending to replace family or clinicians. Below are verified examples, with what they actually do and where they’re used.
PARO (Japan): a therapeutic seal for calm and connection.
PARO looks like a baby harp seal; it blinks, moves, and makes soft sounds when you stroke it. Under the fur, PARO has five kinds of sensors—tactile, light, sound, temperature, and posture—so it can respond to touch and voice and change behaviour accordingly. Peer-reviewed and clinical reports associate PARO use in elder care with reduced stress and agitation and increased engagement, especially for people living with dementia.
ElliQ (Israel/US): a proactive AI companion for older adults at home.
In partnership with Intuition Robotics, the New York State Office for the Aging (NYSOFA) introduced ElliQ to hundreds of older adults as part of a statewide initiative aimed at reducing social isolation. In its 2023 outcomes report, NYSOFA documented a remarkable 95% reduction in user loneliness and consistently high engagement, with older adults interacting with ElliQ more than 30 times per day, six days a week. The robot is designed as a table-top companion that feels like a presence—ready to chat naturally, set reminders, encourage healthy routines, and connect users with family via video calls.
Lovot (Japan): a “be-loved” home robot for warmth and companionship.
Created by Groove X, Lovot is designed not to perform chores but to provide emotional comfort. It approaches people, seeks cuddles, and maintains eye contact—behaviours intended to feel more like affection than automation. According to Groove X’s official materials, its “Emotional Robotics” technology is deliberately engineered to “stir feelings” and evoke warmth through Lovot’s look, feel, and expressive behaviours.
Independent research supports this framing. A peer-reviewed study titled “Improving the Social Well-Being of Single Older Adults Using the LOVOT Social Robot: Qualitative Phenomenological Study” found that interacting with Lovot improved the reported social well-being of single older adults, underscoring its role as a genuine emotional presence rather than just another gadget.
Qoobo / Petit Qoobo (Japan): a tailed cushion for tactile comfort.
Developed by Tokyo-based Yukai Engineering, Qoobo is a soft, furry cushion with a lifelike tail that responds to touch. Stroke it gently and the tail wags slowly; pat it more firmly and it swings with playful energy. The effect is simple but powerful—a tactile form of comfort for people who may want the soothing presence of a pet but cannot keep one due to allergies, housing restrictions, or lifestyle. First unveiled in 2017, Qoobo has since evolved into the more compact Petit Qoobo, expanding its reach as an accessible stress-relief companion.
Evidence from real-world programs shows promising outcomes, though results vary depending on the setting and how the robots are introduced. In ElliQ initiative, the Office for the Aging reported both a decline in loneliness among participating older adults and consistently high levels of daily interaction. Similarly, studies of PARO in nursing homes have documented its ability to soothe agitation and encourage engagement among people living with dementia. These findings underline an important point: the benefits of social robots depend not only on the technology itself but also on how caregivers and users integrate them into everyday routines.
While much of the evidence comes from elder care and therapeutic contexts, social robots are not confined to private settings. Increasingly, they are appearing in public-facing roles—as bank greeters, airport concierges, and hospital companion. Let’s turn to these customer service applications to see what has already been piloted and how people are responding.
Can social robots really work in customer service and public spaces?
Short answer: yes—when the use case is narrow and the environment is prepared. Banks, airports, hotels and hospitals have all piloted customer service robots to greet visitors, answer questions, or guide people to services. Here’s what’s been actually deployed:
Pepper (Japan → global): greeter and guide.
SoftBank’s Pepper is the familiar, wide-eyed humanoid robot with a touchscreen. The company positions Pepper as a social humanoid that can recognise faces and basic emotions and assist visitors. In the United States, HSBC publicly launched Pepper at its Fifth Avenue flagship in 2018 to greet customers and answer common questions—an early, well-documented bank deployment. SoftBank also reports thousands of Peppers used across companies worldwide for hospitality and information roles.
Huggable (United States): a robotic teddy bear for children’s hospitals.
Developed by the MIT Media Lab and trialed at Boston Children’s Hospital, Huggable is a plush teddy bear robot designed to interact naturally with children. In clinical studies, Huggable was shown to reduce stress and anxiety during hospital stays, support positive mood, and make children more willing to engage with caregivers. The trials compared interactions with Huggable to video chat and plush toys, finding that the robotic bear offered unique benefits in comfort and engagement.
Furhat (Sweden → Germany/Japan): multilingual airport concierge.
Developed by Sweden’s Furhat Robotics, the Furhat robot uses a rear-projected animated face on a head-shaped mask, allowing natural eye contact, expressions, and customizable personas. At Frankfurt Airport, Furhat was deployed in collaboration with Deutsche Bahn (DB Systel) and Fraport under the name FRAnny. During the trial, FRAnny acted as a multilingual information assistant, answering passenger questions about gates, directions, and airport services such as Wi-Fi—an officially documented pilot by the airport and rail operators.
It’s one thing for a robot to greet someone, but quite another to hold a meaningful conversation or pick up on subtle emotions. What sets today’s social robots apart from those of a decade ago is the strength of their AI stack—combining speech and vision systems with the latest generation of large language models.
What technology makes social robots feel “alive”?
Several building blocks now combine to make AI companions and customer service robots feel more fluid, less canned, and genuinely responsive.
Conversational AI: Modern robots rely on speech recognition and Natural Language Processing (NLP). The latest step forward is the integration of large language models (LLMs), which allow richer, multi-session dialogue. For example, the Nadine platform has been described in academic work as an LLM-driven social robot, blending memory and affect models to make conversations feel more lifelike.
Perception: Cameras and microphones detect faces, voices, and even coarse emotional states from expressions or tone. This lets robots adapt—for example, slowing down if you look confused. Reviews in human–robot interaction show that embodiment plus perception changes user behaviour compared with screen-only agents.
Embodiment and non-verbal cues: Simple signals like eye contact, head tilts, or a wagging tail raise the sense of presence. Devices like PARO and Qoobo rely heavily on these tactile cues to soothe and connect.
Adaptive behaviour: Companions like ElliQ are designed to be proactive—gently prompting users to follow healthy routines while gradually learning their preferences. Data from the New York State programme showed that older adults not only interacted with ElliQ frequently each day but also reported a noticeable reduction in feelings of loneliness when the robot became part of their daily lives.
As robots become more conversational, perceptive, and proactive, the stakes rise. That leads to the next question: what responsibilities come with putting such lifelike companions into homes, clinics, and public spaces?
What are the key ethical issues of social robots?
The ethics of social robots aren’t abstract; they appear in real living rooms, care homes, and busy lobbies. Reviews and field studies consistently highlight four key issues:
Over-attachment and dependency: Companion robots can comfort, but they risk over-reliance. Care teams stress the need to set expectations—robots simulate empathy, they don’t feel.
Deception and transparency: Human-like dialogue, faces, or pet-like behaviour can encourage over-anthropomorphism. With LLM-driven systems, there’s also a risk of overstatement, so clearer messaging and guardrails are critical.
Privacy and data protection: Cameras, microphones, and cloud services create sensitive data trails. Analyses recommend privacy-by-design and clear consent in eldercare and home use.
Equity and inclusion: Value-sensitive design research shows rollouts should account for cultural context, accessibility, and marginalised groups—so robots don’t exclude the very people who could benefit most.
These ethical concerns remind us that social robots are not just gadgets but participants in human environments. Setting clear guardrails—around dependency, transparency, privacy, and inclusion—is essential to ensure they enhance rather than undermine social life. With those responsibilities in mind, the next step is to ask: where is the field heading globally, and what trends are shaping the future of social robots?
What is the future of social robots around the world?
The next wave of social robotics won’t just repeat existing pilots—it will push into new frontiers of everyday life. Several future directions are already taking shape:
From niche to mainstream living spaces: Social robots are moving from eldercare facilities and banks into ordinary households, marketed as lifestyle companions. Devices like Sharp’s Poketomo, set to launch in 2025, point to a future where people casually buy a “pocket-sized friend” much like they would a smart speaker.
Deeper emotional intelligence: With advances in large language models and affective computing, the next generation of robots will be able to sustain conversations across weeks or months, remember personal context, and respond more sensitively to emotional cues. This could make them feel less like gadgets and more like long-term companions.
Integration into public infrastructure: Expect more social robots in airports, hospitals, and transport hubs, not just as novelties but as permanent fixtures that handle multilingual assistance, ticketing, or navigation.
Personalisation at scale: As hardware costs fall, robots will become cheaper and more customisable, allowing users to “tune” their companion—choosing a playful pet-like design, a professional concierge, or a calming presence for stress relief.
Policy and regulation catch-up: Governments will likely face pressure to create ethical standards for transparency, data use, and safety in social robotics. Just as we now have frameworks for medical devices and AI in healthcare, social robots will need similar guardrails to build public trust.
Blending physical and virtual: Looking further out, social robots may bridge the physical and digital worlds—acting as an embodied avatar for remote family members, healthcare providers, or even colleagues, giving a human touch to video calls or telepresence.
Over the next decade, the big question will be whether people accept robots as everyday companions—not just tools or novelties. Success will depend on striking the right balance: combining powerful AI with designs that respect privacy, foster trust, and above all, enhance human relationships rather than replace them.
.png)
